It is vitally important, not just in elections but also in life overall, to recognize that not all possibilities are equally likely and to use available data to evaluate them.

For example, it is possible that a white rhinoceros might come down the hall and join me for a spot of tea and biscuits. I can, however, safely say that that is almost definitely not going to happen, based upon these data points:

• I do not keep any white rhinoceros in my home.
• White rhinoceros are not native to this continent and are classified as near-threatened (a major improvement from their previous near-extinction).
• White rhinoceros average between 1700kg and 2300kg. My home was not designed and built to support that kind of moving mass. Any white rhinoceros walking down my hall would likely fall through the floor.
• White rhinoceros, while generally docile and gregarious, are not known for enjoying tea and biscuits.

Yes, this is a exaggerated caricature. However, data with even more clarity is available surrounding election outcomes. Betting on a possible outcome without any evidence to suggest that it is likely is foolish and betting on an outcome that one knows is unlikely, while increasing the likelihood of an outcome known to be harmful to vulnerable people is detestable.